Monthly Archives: November 2015

SpatialOS: What If We Had A Digital City?

In this video from Slush 2015 the CEO of tech company Improbable, Herman Narula, introduces the company’s ground breaking SpatialOS by asking the question ‘What if we had a digital city?’  Its value is the possibility it provides for seeing into the future of the city and answering the types of questions that start ‘What if?’ In this way the concerns motivating the creation of SpatialOS coincide with those of my own PhD in Near Real Time Urban Data Spaces here at CASA.

In his presentation Narula considers the impact that the introduction of self-driving autonomous vehicles might have on our cities. We might expect them to change patterns of usage on transport networks by making journeys more efficient, and this could have further implications for land uses associated with related phenomena such as parking. This in turn suggests an impact on economic activity which, in turn, could have implications for transport related crime and policing. We can guess what the impacts might be but what will this really look like?

The motivation behind SpatialOS is the ability to understand the consequences of complexity. In complex adaptive systems such as cities the interactions between different components of the system are non-linear. Contrasted with the relationships in linear systems where effects will be proportional to and more directly related to their causes, the relationships in non-linear systems will be disproportionate, indirect and, therefore, extremely difficult to predict.

While large amounts of data about cities can now be gathered in near real time for the purpose of up to the minute analysis, this does not provide any guarantee that the currently observed patterns of behavior will persist into the future. This limits our ability to answer questions of the kind ‘What If?’ What if we add a new station to the tube network? What if we route the trains differently? What if we build a new district? What will the impact of a new technology be? SpatialOS offers to help answer these questions by enabling us to digitally recreate and simulate the entire system.

The challenge Narula identifies is that of achieving a form of ‘Strong Simulation’ which leverages sufficient computational power that is fully dedicated to the process of simulation. This is contrasted with ‘Weak Simulation’ which Narula characterised in a previous presentation as the necessity of sharing computation between simulation and other processes such as rendering. With SpatialOS Strong Simulation is achieved through the distribution of computational load across servers in the cloud.

This is potentially ground breaking technology for data scientists but also has amazing potential for the creation of virtual environments and video gaming. The presentation ends with a presentation from Bossa Studios whose game Worlds Adrift uses SpatialOS to provide a massive mutliplayer online gaming environment.

What makes the game special is the way in which each of the individual objects in the game environment are indefinitely persistent and affected by physics in real time. Already simulating over four million objects at the current stage of development the potential of SpatialOS for both gaming and scientific simulation is massive!

 

VUCITY: Wagstaffs 3D Model of London Ten Months On

Back in February I posted the first video of VUCITY by Wagstaffs Design. Ten months on the functionality of the product has improved considerably. Aimed at high end developers, planners and architects their 3D model visualises 80 square kilometres of London and incorporates high detail contextual models provided by Wagstaffs partners Vertex Modelling.

In addition to interactive navigation the model also provides tools for rights to light and sunlight studies, London View Management Framework (LMVF) visualisation for protected views, search for existing, planned and consented development, transport information and integration of live data overlays. At the recent MIPIM expo in London VUCITY also appeared with further functionality for intuitive navigation through gesture control.

In a recent feature on London Live Wagstaffs director Jason Hawthorn explains how VUCITY has been created with the aid of gaming technology, in this case Unity game engine, and outlines the companies future plans. VUCITY remains one to watch closely!

GSS 2015: Presenting INSIGHT on behalf of CASA

GSS

This past week I’ve had the good fortune to be in Genoa in northern Italy participating in the Global Systems Science (GSS) conference 2015. where I represented CASA on behalf Dr Vassilis Zachariadis who was unable to attend. This year the conference was jointly organised by the Genoa Festival of Science. The focus of the conference is the way in which global issues impacting the globe can be addressed using data analysis and systems science to support to coordinate strategies through policy:

Philosophers have been reminding us for millennia that everything in the universe is connected to everything else in one enormous complex system. While this has always been true, until recently the propagation of states of the many sub-systems was so slow that they appeared to be disconnected. The advent of globalisation, the internet, and instantaneous propagation of system state have outrun human ability to understand the complexity of modern, global systems. In parallel, our ability to capture information about events in the real-world, to communicate, to store, and to analyse such information has made extraordinary progress. We are approaching a state where almost everything that determines how we live is capable of being observed and analysed. Global systems science (GSS) is a policy-related scientific program dedicated to research in applying systems science and large scale data analysis and models to global challenges including climate change, pandemics, sustainable growth, energy sufficiency, financial crisis, urbanisation and global conflict. These challenges are all ‘global’ and ‘borderless’, i.e. they are shared worldwide and they tightly connect policies across different sectors. They necessitate joint action and coordination across various networks of actors with different interests. The conference will elaborate on how systemic thinking, models and data can help address such coordination problems in policy, society, and economy.

CASA_INSIGHT_GSS

CASA’s contribution to the conference involved the demonstration of visualisation outputs relating to research conducted as part of the pan European project INSIGHT (Innovative Policy Modelling and Governance Tools for Sustainable Post-Crisis Urban Development). This project seeks to investigate how information and communication technology (ICT), data science, and complexity theory can help European cities formulate policies for for sustainable urban development and economic recovery in the event of crisis. The objectives of the project are:

1. to investigate how data from multiple distributed sources available in the context of the open data, the big data and the smart city movements, can be managed, analysed and visualised to understand urban development patterns;
2. to apply these data mining functionalities to characterise the drivers of the spatial distribution of activities in European cities, focusing on the retail, housing, and public services sectors, and paying special attention to the impact of the current economic crisis;
3. to develop enhanced spatial interaction and location models for retail, housing, and public services;
4. to integrate the new theoretical models into state-of-the-art simulation tools, in order to develop enhanced decision support systems able to provide scientific evidence in support of policy options for post-crisis urban development;
5. to develop innovative visualisation tools to enable stakeholder interaction with the new urban simulation and decision support tools and facilitate the analysis and interpretation of the simulation outcomes;
6. to develop methodological procedures for the use of the tools in policy design processes, and evaluate and demonstrate the capabilities of the tools through four case studies carried out in cooperation with the cities of Barcelona, Madrid, London, and Rotterdam.

The INSIGHT project which commenced in October 2013 is funded by the European Union’s Seventh Framework Programme and will conclude in 2016. Alongside CASA the project is run by a consortium including the Technical University of Madrid (UPM), Nommon Solutions and Technologies, the Technical University of Eindhoven (TU/e), the Institute for Cross-Disciplinary Physics and Complex Systems (IFISC) at the University of the Balearic Islands (UIB), and Barcelona City Council.

A number of CASA visualisations relating to the themes of the project were presented over the course of the conference. At the pedestrian level the following animation by Vassilis demonstrates the output of an agent based model involving the dynamic assignment of pedestrian route choice in response to crowding. The colour of each pedestrian reflects the deviation from their preferred walking speed:

  • Red = high deviation
  • Green = Low deviation

More details on the used methods can be found in the following paper:

Zachariadis, V., Amos, J., Kohn, B. (2009) Simulating pedestrian route choice behaviour under transient traffic conditions.

Looking at the urban scale Vassilis’ second visualisation uses a cross-cluster analysis of geolocated tweets in London through the course of a day. In this way it becomes possible to see how urban activities vary in different locations across the city at different times of the day. The different colours reflect estimated activity based on data concerning the known uses of premises associated with those locations:

  • Yellow = Retail
  • Red = Offices
  • Cyan = Rail
  • Green = Leisure

Further details of the analysis behind the visualisation can be found here.

By way of contrast Vassilis also provided an earlier visualisation demonstrating the origins and destinations of journeys by taxis equipped with GPS trackers in the San Francisco Bay Area over a period of 24 hours. In this visualisation green dots represent pickups while red dots represent drop offs:

When watched carefully the visualisation shows how certain areas shift from being more predominantly collectors or distributors of taxi trips throughout the day. In this way it suggests a ‘tempo of the city’ as Vassilis notes in his original blogpost here.

Now lecturing at King’s, Dr Jonathan Reades’ visualisation for CASA of Oyster card data entitled ‘Pulse of the City’ further develops this theme of tempo in the dynamics of passenger flows across London’s tube network:

The next visualisation by CASA researcher Joan Serras presents the output of a model for activity-based transport demand created by Joan, Melanie Bosredon, Vassilis Zachariadis, Camilo Vargas-Ruiz, Thibaut Dubernet and Mike Batty. Using MATSim software and TFL data they simulated activity across a typical working day in London. Each person or agent within the model is plotted performing certain activities:

  • Dark Blue = Home
  • Light Blue = Work
  • Yellow = Shop
  • Green = Education
  • Pink = Leisure

The model was developed by CASA under the EUNOIA research project for Evolutive User-centric Networks for Intraurban Accessibility. The output of the model was visualised using Via software by provided by Senozon. Further details are available here.

On a national scale Joan Serras’ visualisation of Public Transport flows across the UK for a typical day in 2009 was also displayed. Presenting data on train, coach, tram, tube, ferry and available air trips for Scotland, the visualisation demonstrates the complexity of the networks and indicates the distinct transport geographies of particular regions.

Further information on the visualisation can be found here.

I also had the opportunity to display sneak previews of a visualisation presenting the application of land use modelling to Dubai along with an interactive model of the UK being developed by CASA and the Future Cities Catapult.

During the conference I had fantastic opportunities to attend plenary sessions on complexity by Geoffrey West from the Santa Fe Institute and on ‘The Promise of Urban Science’ by Steven E. Koonin of New York University. I also had a great time exploring Genoa’s unique topography and the labyrinthine knot of its old town. Many thanks to CASA and Vassilis for making the trip possible.

Finally, my thanks also to Luca Piavano and Francisco Oostrom from CeDint at UPM who joined me at the INSIGHT stand to present their own great work. I owe them both a pint next time they are in London.